THE INFLUENCE OF HEART RATE ON LEFT VENTRICULAR VOLUME IN DOGS

THE INFLUENCE OF HEART RATE ON LEFT

VENTRICULAR VOLUME IN DOGS

J. David Bristow, … , Fredric Mintz, Elliot Rapaport

J Clin Invest.

1963;

42(5)

:649-655.

https://doi.org/10.1172/JCI104755

.

Research Article

Find the latest version:

Vol. 42, No. 5, 1963

THE INFLUENCE OF HEART RATE ON LEFT VENTRICULAR

VOLUME IN DOGS

*

By J. DAVID BRISTOW,t RICHARD E.

FERGUSON,$

FREDRIC MINTZ,§ AND ELLIOT RAPAPORT

(From the Cardiovascular Research Institute and the Department of Medicine, University of

CaliforniaSchool of Medicine, and the Cardiopulmnonary Laboratory, San Francisco, General Hospital, San Francisco, Calif.)

(Submitted for publication August 10, 1962; accepted January 17, 1963)

The effects of changes

in heart rate must be

understood before ventricular performance can be

evaluated in studies of the response to drugs,

in-jury, or abnormal circulatory states.

This

re-port describes the effects of sudden changes in

heart rate on the stroke, systolic, and

end-diastolic volume of the canine left ventricle.

The

ventricular volumes were measured by

thermodi-lution, an indicator dilution method in which

in-jected cold blood serves as the indicator (1).

METHODS

There were 160 experimental periods in 11 mongrel dogs that were anesthetized with a chloralose-urethane mixture. Aortic thermodilution curves were obtained after rapid injection of cooled autologous blood into the left ventricle. A 4F catheter with an ultrasmall bead thermistor secured to its end 1 was used to record blood temperature. It was introduced into a carotid artery and advanced so that its tip was just above the aortic valve. The time constant ofthis assembly was 0.12 sec-ond when tested in slowly flowing water. A 7F catheter with multiple side holes and a closed end was manipu-lated into the left ventricle from a femoral artery for

injection purposes and to record left ventricular pressure with a strain gauge. A bead thermistor was secured

within its lumen near the tip to monitor the temperature of theinjectedbloodimmediatelybefore it entered the left ventricle. This assembly had a time constant of 0.05 second in flowing water.

*_{Supported by U. S. Public Health Service grants}

H-6285 and H-5817 and a grant from the San Mateo

County Heart Association.

t

Work

performed during

the tenure of a special

fel-lowshipof theU. S. Public Health Service

(HF-13,456).

1

\Vork

performed

during

the tenure of a

_postdoctoral

fellowship of the U. S. Public Health Service (HF-14,-505).

§ Work performed during the tenure of a postdoctoral

fellowship of the U. S. Public Health Service

(HF-14,451).

1 Victory Engineering Corp., Union, N. J.

Injections were made with a metal syringe driven by compressed air that introduced a known amount, usually

3 to 5 ml, of cooled, heparinized, autologous blood into the left ventricle in less than 0.5 second. The mean

tem-perature of the inj ected blood was determined by

plani-metry of the recorded temperature curve during injec-tion. Stroke volume (SV), end-systolic volume (ESV),

and end-diastolic volume (EDV) were calculated from the resultant aortic thermodilution curve sensed by the aortic thermistor catheter. The formulas used to calcu-late these volumes were:

SV(ml) =

Vi(Tb-

) 1)

where Vi = _volume of injected cooled blood in ml, Tb =left ventricular blood temperature before injection,

T1

= mean temperature of injected blood, just before it entered the left ventricle, and 2;(ATb) = the sum of the differences between base-line aortic temperature and that resulting in the aorta from each systole, as measured at

end-diastole;

2) ESV/EDV=

AT,+,

ATn'

where AT,, and

AT,,+,

are differences between base-line aortic temperature and that at beats n and n + 1,

re-spectively, measured at end-diastole from theexponential step-function of the aortic thermodilution curve;

3) EDV(ml) = SVES V; EDV

ESV(ml) = _EDV- SV; 4)

and

5)

Cardiacoutput

(L/min)

= (heartrate)(S V)

The details of this method have been published previously

(1)

-One to five determinations of left ventricular volumes,

left ventricular pressure, and heart rate were made dur-ing each control and experimental period. Various heart

rates faster than control were produced with a 6F elec-trodecatheterpassedupafemoral vein andplaced fluoro-scopically against the right atrial appendage. The other electrode was on the right chest wall. Pacing was

J. D. BRISTOW, R. E. FERGUSON, F. MINTZ, AND E. RAPAPORT

achieved with a stimulus of 1 to 5 v lasting 10 msec as

provided by a Grass model S4DR stimulator. The

dif-ferent rates at which the hearts were paced were

ran-domly chosen, and control measurements were repeated

after eachnew ratechange. The effects of pacing per se

were studied in six experimental periods in different ani-mals when the pacing rate was within 2 beats per min-uteofthe heart rate presentat the onsetofpacing. The

average ESV/EDV ratio (in per cent) for these six

pairs of observations was 73% before and 72%o after

pacing was established. Stroke volume was an average

of 16.3 ml both before and with pacing.

Heart rates slower than control were produced by

stimulating the peripheral end of the right vagus nerve

which had been severed in the neck. Usually, an

elec-trical pulse of 10c per second and 1 ma produced

brady-cardia. If the heart rate became irregular at the slow

rate, or ifventricular standstill occurred, a constant slow

rate was obtained by simultaneously pacing the right

atrial electrode catheter with a rate somewhat faster than that resulting from vagal stimulation alone. Only curves that were associated with a regular rhythm were analyzed..

The twobasic measurements made with the

thermodi-lution curves were SV and the ESV/EDV ratio, as shown in Formulas 1 and 2. The ratio ESV/EDV is

probably the more accurate of the two, since it is not dependent upon an exact amount of injected indicator. It represents the degree of emptying characteristic of the ventricle at the time, and is the proportion of the

EDV that remains in the ventricle at the end of systole.

Inorderto assessthe reproducibilityof the results in this study, 20 experimental periods were chosen randomly

fromseveral animals. During each,twoor three

thermo-dilution curves had been obtained, with a total of 44

curves. The variations of SV and ESV/EDV for re-peat determinations within an experimental period were expressedas a percentagedifference from the average for that period, regardless of sign. The mean variation for

ESV/EDV from the average for any experimental

pe-riod was 3.4 3.3% (mean 1 SD). This agrees closely with our previous observations on the

reproduci-bility of themethod (1). Stroke volume variations from

the average within an experimental period were 8.2 +

7.1%.

Inearly experiments with thermodilution, wecompared

the cardiac outputs obtained by Evans blue dye and ther-modilution following left ventricular injection of cold Evans blue dye. The ratio of outputs calculated from

thermodilution curves to dye dilution outputs averaged 1.09 (SD, .12). We believe thatthis reasonable standard deviation has been further reduced by subsequent

im-provements in our technique, including the substitution of the animal's own cooled blood for saline as the injection (eliminating the problems of specific heat capacity and specific gravity) and more accurate registration of the injection temperature.

TheESV/EDV ratio and all of the volume calculations depend on an exponential washout of indicator from the

left ventricle. In routine analysis of the thermodilution curves, the series of beat to beat downslope temperature

ratios, T3/T2, T4/TS,

TI/T4,

and so forth, were averaged to obtain the ESV/EDV ratio for that curve. To test

the variability of the downslope exponential functions we observed, 46 randomly chosen curves from similarly

con-ducted previous experiments were analyzed in more

de-tail. The variation of each T,+ITn ratio in the down-slope was compared to the average of such ratios for a

BLE I

Originalcontrolmeasurementsin all animals*

Left

End- End- ventricular

Heart Cardiac Stroke diastolic systolic systolic

Dog Weight rate output volume volume volume ESV/EDV pressure

kg beats/min Limin ml ml ml % mmHg

1 17.7 135 1.37 9.9 39.6 29.7 74 166

2 18.6 167 1.85 11.3 43.5 32.2 74 184

3 20.5 199 2.87 14.6 37.4 22.8 61 185

4 15.9 151 1.57 10.7 34.0 23.3 66 213

5 19.1 121 1.22 10.0 47.6 37.6 79 177

6 34.1 152 1.28 8.4 26.9 18.6 69 153

7 29.5 157 3.40 21.7 67.8 46.2 68 203

8 29.5 112 2.19 19.4 59.8 40.4 68 161

9 21.8 86 3.41 39.7 110.5 70.8 64 189

10 23.6 126 1.08 8.7 35.5 26.9 76 165

11 24.1 112 1.68 14.6 51.7 37.2 72 213

Mean 23.1 138 1.99 15.4 50.4 35.1 70.1 183

SD 31.3 0.86 9.2 23.2 14.5 5.4 20.7

* Each value listed for each animal isanaverage of all measurements during the first control period of the experiment. Heart ratewasmeasured from records of left ventricular pressureaswell asthethermodilution curves. Thus theproduct ofrateandstrokevolumeaslisted above maynotexactlyequal the cardiac output in the table. Inexperiments 1 and 4,

severalthermodilutioncurves wereobtained forestimatingthe ESV/EDVratiobesides those usedtomeasurethe actual volumes.

TABLE II

Effectofchangesin heart rate on cardiac output and left ventricular volumes

Heart Observa- End-diastolic End-systolic

rate* tions Cardiac output Stroke volume volume volume

°%0 no. % % % %

40- 59 16 82.2 ± 56.5 158.5 i 99.0 133.3 i 81.9 122.1 i 75.6 60- 79 10 86.4 i 31.8 132.2 i 48.8 99.8 i 29.0 86.4 i 24.5

80- 99 22 93.0 i 29.8 100.5 :1 29.0 93.9 41 30.0 90.4:1 35.1 100-119 46 106.7 i 30.2 99.0 i 27.0 93.7 4 23.4 91.3 i 24.7 120-139 20 117.3 i 29.2 92.8 4± 22.5 90.6 i 23.6 90.1 i 27.3

140-159 12 103.3 ± 25.5 64.4 ± 21.6 65.3 i21.6 65.8± 23.0

160-179 10 108.4± 25.0 64.4 i: 14.5 63.7 i 15.3 64.1 ± 17.3 180-260 13 94.7 ± 29.6 47.7 i 17.0 58.2 i 16.6 63.0 i 18.4

* Inorder to compare different animals and to pool the results for statistical analysis, all values after the original control measurements were expressed as percentages of these original controls. Theseinclude observations during normal rhythm between periods of pacing or induced bradycardia. Means+ 1 SD are listed.

given curve. The difference from the average was

ex-pressedas apercentage. For 156ratios in these 46curves,

the mean deviation fromthe average was4.9 3.8%.

RESULTS

The

results

are

listed in Tables

I to IV.

After

the original

control observations in each animal,

all subsequentmeasurements ofheart rate, cardiac output,

and

ventricular

volumes

were

expressed

as

percentages

of

the original

control values.

The

ESV/EDV ratios

were

always expressed

as

ab-solute values (ESV/EDV

x

100

=

%),

in

order

to

avoid

the confusion which would have resulted

from

the

use

of

a

proportion

of

a

ratio.

There

were no

significant differences between

the values found

at

80

to

99%

and 100

to

119%

of

the

original

rates.

These results occurred for the

most part

during

periods of normal

rhythm

be-tween runs

of

electrical pacing

or

induced

brady-cardia.

For

statistical

comparison

with

faster

and

slower

rates,

the results between 80 and

119%

of the

original heart

rates were

pooled

as a

con-trol

group.

Comparison

with other

rate groups

was

made

by

the

t test.

Significant findings

were

considered

present

when

the

probability value

found

was

less

than 0.05.

Tachycardia above 140%o

of control caused

a

significant

decrease in

SV.

The

SV

at

the fastest

rates

(> 180%

of

control)

fell

to

less than half

of the

original

values.

Bradycardia

below

80%o

of

control

was

associated with

anincreased

SV,

which at

the slowest

rates

increased

by

a mean value of

58.5

%.

The

changes

in

ESV and EDV

paralleled

those

of

SV,

but

were

less marked.

With

extreme

tachycardia

the EDV was

587c

of control. At

the

slowest

rates,

a

mean value for EDV

of

133%

was

found, but

was not shown

to

be

statistically

differ-ent from

control

at the 0.05

probability

level.

Gen-erally

similar

changes

were

found with

the

ESV.

TABLE III

Probability values from comparisonof pooled resultsat rates of80 to119%of theoriginal controls withfasterandslower

rates*

Heart End-

End-rate Cardiac Stroke diastolic systolic (%of output volume volume volume

original) p p p p

40-59 <0.025

60- 79 <0.050

120-139 <0.050

140-159 <0.001 <0.001 <0.005 160-179 <0.001 <0.001 <0.001 180-260 <0.001 <0.001 <0.001

TABLE IV

End-systolic volumetoend-diastolic volume

ratiosfor allexperiments*

Heart

rate Probabilityvalues

(% of Obser- fromcomparisonwith

original) vations ESV/EDV pooled control group

no. % p

40- 59 16 62.647.1 <0.020 60-79 10 62.446.9 <0.050 80- 99 22 65.4±5.8

Pooledcontrolgroup 100-119 57 67.9±6.0J

120-139 20 7 1.8 47.2 <0.010

140-159 12 71.0±5.5 <0.050 160-179 10 74.1±4.2 <0.001 180-260 13 76.4±5.0 <0.001

* Therewas nodifferencebetweenthegroupswhichwere80to99%

and100to119% oftheoriginalcontrolrates. Thesetwogroupswere

pooled for statisticalcomparisonwith otherrategroups. Theoriginal

11 control observationsareincluded. Means± 1SDarelisted in the

J. D. BRISTOW, R. E. FERGUSON, F. MINTZ, AND E. RAPAPORT

It

decreased significantly with

fast rates, but was

not proven

statistically to

have increased with

bradycardia.

The mean ratio of

ESV to EDV

during the first

AT

₀₀

A.

AT

4.3*

control

periods

was

70.1% (SD, 5.4%o)

and

is in

agreement with other control

observations from

our

laboratory for

the dog anesthetized with

chlo-ralose and

urethane

(2). At rates above

120%

of

S-3

-38.0

38.2

k30.4

.

*37.6

*37.8

B. W .1 _I

1

I I

_,1

i

_I.

_IIlII

i

FIG. 1,A AND B. THEEFFECTS OF EXTREME TACHYCARDIA ON THERATIO OF END-SYSTOLIC VOLUME (ESV) TO END-DIASTOLIC VOLUME (EDV). In both

examples the upper curve is the record of the temperature of the injected cooled blood, and a downward deflection indicates a lower temperature. AT is the difference between left ventricular blood temperature before

in-jection and the peak coldness of the injected blood. At the bottom is the aortic thermodilution curve. The temperature scale on the right applies to

it, and an upwarddeflection results from lowered temperature. The height above the base line of each diastolic plateau on the downslope ofthe aortic curve, dividedby the height of the preceding one, equals the ratio of ESV toEDV. A. Withaspontaneousrateof100,theESVwas

53%0

oftheEDV.

B. When pacedat 268 perminute, the

ESV/EDV

was

81%o,

and the stroke

volume

(SV),

therefore, was only 19% of the EDV. 652

AT

1I..

2-55

FIG. 2. VAGAL STIMULATION AND PACING. When the rate was slowed to 100, the ESV/EDV was 64%, a typical value at this rate. The artifact from the electronic pacemaker is shown by the arrow.

control,

this residual fraction increased and

be-came

progressively

larger with more extreme

tachy-cardia

(Figure

1

and Table IV).

The opposite

was

observed with bradycardia (Figure 2). The

mean values for

SV, EDV, and ESV are plotted

in

Figure 3, and

further illustrate

the manner in

which the ESV to EDV ratios changed.

At

the two extremes

of

rate,

cardiac output did

not

change significantly, since opposite changes in

SV counterbalanced the

rate

changes.

With

rates

only

slightly faster than control

(120

to

139%),

cardiac output

rose

17%,

which

was

a

significant

change. This resulted

from

an

increased

rate

and

an

unchanged

SV.

A4J

150

100

0

ost

DISCUSSION

Rushmer noted that the change in diameter

of

the dog's left ventricle caused by each stroke

out-put

decreased

as

the heart rate was increased

by

right atrial electrode pacing (3).

Warner

and

Toronto

(4) and Miller and his colleagues (5)

also

demonstrated

a

progressive

fall in

SV

as

pacing increased the

ventricular rate

in

dogs

with

surgically produced complete

heart

block.

Our

results agree

with these observations

at

the

ex-tremes

of

rate

change.

Between 80 and

140%o

of

control rates, however,

there was no

significant

change

in

SV.

Consequently,

cardiac output

rose

240

FIG. 3. THEMEANVALUES FORSV, EDV, ANDESV AREPLOTTED IN THE

MIDDLE OFEACH RATE GROUP. SV shows the greatestchange over the entire rangeof rates, and ESV the least.

60

S0

100 120 140 160 180

J. D. BRISTOW, R. E. FERGUSON, F. MINTZ, AND E. RAPAPORT

200

175

150~

40

14

01

125

100

751

SO f

25-*

-*

0

.0

1

25 50 75 100 125 150 175 % OF CONTROL EDV

200 225 250

FIG. 4. THE RELATION BETWEEN ESV, AND EDV IS ILLUSTRATED. The

plot as percentage of control volumes suggests that these two volumes

changed in proportion to eachother. Individual animals, however, had les-ser changes in ESV than in EDVat the extremes of rate change.

slightly but

significantly with rates 20 to 40{%o

above control levels.

These observations provide an interesting

con-trast with

studies which have shown that SV does

not

fall

with the tachycardia of exercise, even at

rates

faster than those in our study (6-8). The

exercise response of stroke volume to tachycardia

differs from the results of electrical pacing,

al-though

the EDV becomes smaller in both

situa-tions. Diastolic filling time is shorter in both

cir-cumstances

and cannot account for the difference.

If

arterial blood pressure has

not

decreased, the

maintenance of

a

normal SV,

despite

a

smaller

ED V during

exercise, implies

that

there

are

mechanisms

to

increase

ventricular

emptying

as

well

as to

maintain

an

adequate

venus return.

With

fast

pacing

in

our

experiments, opposite

ef-fects

were

observed,

with decreased

stroke volume

and proportionately decreased ventricular

empty-ing

(higher ESV/EDV).

Previous studies have

shown

a

generally

linear

relationship

between

the

ESV and the EDV

(2,

9,

10).

This

was

observed

again

in

this

study

over

the wide range of heart

rates

produced (Figure

4). A small ESV

at

fast

rates was

usually

asso-ciated with

a

small

EDV, whereas

a

larger

ESV

was

associated

with a

larger

EDV

at

slower

rates.

Since the plot of all of the experimental values in

Figure 4

represents percentage

of

change

from

each individual

animal's own control values, the

distribution appears to pass through the origin.

However,

a

linear regression equation of the

ab-solute

values

does not pass through the origin.

This is in keeping with the fact that values from

individual

animals

consistently indicated a higher

ESV/EDV with tachycardia and the opposite

with

bradycardia.

Thus,

ED

V fell

more

than

ESV

at

fast rates, and the ventricle emptied

ac-tually

and

proportionately less.

At

slow rates,

ventricular

emptying

was

proportionately

greater,

despite

a

larger ESV, since

a

greater part

of

the

EDV

was

ejected with systole.

These

findings

support a-concept

that the

left ventricle does

not

function with a

relatively constant end-systolic

or

INFLUENCE OF HEART RATE ON LEFT VENTRICULAR VOLUME IN DOGS

degree.

Stroke volume was influenced the most,

and the ESV the least by a given alteration in

rate.

SUM MARY

The

effects of sudden changes in heart rate on

left ventricular volumes were studied in dogs.

Fast rates were produced by electrical pacing and

slow

rates

by efferent stimulation of the right

vagus nerve, with and without slow pacing.

The

left ventricular stroke and the end-systolic and

end-diastolic volumes were measured by an

indi-cator

dilution method in which temperature was

the indicator.

Cooled blood was injected into the

ventricle, and aortic thermodilution curves were

recorded

by

a

thermistor catheter.

The left ventricular volumes decreased during

moderate

tachycardia.

At the fastest

rates

pro-duced, the fall in stroke volume

was

proportion-ately greater than that in systolic and

end-diastolic volumes.

At

slow rates, stroke volume

increased

proportionately

more

than

the other

two

volumes.

The findings indicate that the left ventricle

in

these

experiments did

not

function with

a

constant

end-systolic

or

residual volume.

However, the

directionally similar changes in

end-systolic

vol-ume were

of lesser

proportion

than the stroke

vol-ume

alterations.

REFERENCES

1. Rapaport, E., B. D. Wiegand, and J. D. Bristow. Estimation of left ventricular residual volume in the dog by a thermodilution method. Circulat. Res. 1962, 11, 803.

2. Bristow, J. D., R. E. Ferguson, F. Mintz, and E. Rapaport. Thermodilution studies of ventricular

volume changes due to isoproterenol and bleeding

J. appl. Physiol. 1963, 18, 129.

3. Rushmer, R. F. Constancy of stroke volume in

ven-tricular responses to exertion. Amer. J. Physiol. 1959, 196, 745.

4. Warner, H. R., and A. F. Toronto. Regulation of cardiac output through stroke volume. Circulat.

Res. 1960, 8, 549.

5. Miller, D. E., W. L. Gleason, R. E. Whalen, J. J.

Morris, Jr., and H. D. McIntosh. Effect of ven-tricular rate on the cardiac output in thedog with chronic heart block. Circulat. Res. 1962, 10, 658.

6. Rushmer, R. F.,

0.

Smith, and D. Franklin. Mech-anisms of cardiac control in exercise. Circulat.

Res.

1959,

7, 602.

7. Wang, Y., R. J. Marshall, and J. T. Shepherd.

Stroke volume in the dog during graded exercise.

Circulat. Res. 1960, 8, 558.

8. Chapman, C. B., 0. Baker, and J. H. Mitchell. Left ventricular function at rest and during exercise.

J. clin. Invest.

1959,

38, 1202.

9. Holt, J. P. Effect of plethora and hemorrhage on

left ventricular volume and pressure. Circulat. Res. 1957, 5, 273.